Model Follower.

Up to date

Rebuilt
2026-10-25 06:54 UTC
Release radar
117 on watch
Hub snapshot
2026-10-25
Re-checked
daily · every number sourced and dated

Models › gemma-2 family

gemma-2-27b

Hugging Face ↗ Follow changes

google Gemma Terms of Use out 2 years ago

27.2B parameters, text in, text out, 8K tokens of context. 2 providers serve it from $0.65 per million input tokens.

Verified 2026-10-25 Rank #622 in LLMs · score 73.5 method v1 Sources 3 Downloads 30d 5.1k · #543 of 619 tracked

Can I run it?16GB at INT4Estimatefits a 24 GB cardmethod
What does it cost?$0.65/1M incheapest of 1NextBitint42026-10-24
Can I use it?Gemma Terms of UseGoogle's terms + prohibited-use policy; must pass terms downstreamlicence file2026-10-24
Do people use it?5.1kdownloads, 30 days#543 of 619 we track · below the median+0 likes, downloads −3% this week211 likesHugging Face2026-10-24

Worth reading

Picked by rule from 16 tracked pieces: the lab's own post, the most shared, a field report, the latest.

Most discussed · 328 HN pts

Gemma 2: Improving Open Language Models at a Practical Size [pdf]

Google DeepMind releases Gemma 2, a family of open language models ranging from 2 billion to 27 billion parameters.

  • The 2B and 9B models are trained with knowledge distillation instead of next token prediction.
  • Gemma 2 models use interleaved local sliding window and global attention layers in every other layer.
storage.googleapis.com 2024-06-27 328 HN pts · 172 comments 9.9k words about the gemma 2 family family

16 best of 16 tracked

The 2 picks above, plus these 14. Top is the radar's score: discussion, source authority and the lab's own posts, fading with age.

  1. ArticleGemma 2B 60 tk/s WebGPU on M1mediapipe-studio.webapps.google.com · 2024-03-07 · 30 HN pts · official
  2. ArticleGemma explained: What's new in Gemma 2developers.googleblog.com · 2024-08-22 · 14 HN pts · official
  3. DiscussionNew FOSS All-Local AI with Ollama, Gemma 2, and Fabricnews.ycombinator.com · 2024-07-01 · 33 HN pts
  4. DiscussionRun Google Gemma 2 2B 100% Locallynews.ycombinator.com · 2024-07-31
  5. Article2x faster Gemma 2 finetuning and 63% less VRAMunsloth.ai · 2024-07-04 · 3 HN pts
  6. ArticleColab Notebook for finetuning Gemma 2 2B IT modelcolab.research.google.com · 2024-08-03 · official
  7. ArticleGoogle releases Gemma 2 2B, ShieldGemma and Gemma Scopehuggingface.co · 2024-08-01
  8. CodeShow HN: Finetune a Gemma 2B model for codegengithub.com · 2024-02-28 · 5 HN pts
  9. CodeGemma 2B with 10M context lengthgithub.com · 2024-05-09 · 6 HN pts
  10. ArticleGemma 2 on AWS Lambda with Llamafileunremarkable.ai · 2024-07-06 · 10 HN pts
  11. DiscussionGemma 2B: Scores better than GPT 3.5 Turbotwitter.com · 2024-08-01 · 8 HN pts
  12. CodeOllama v0.1.47 release. Added support for Google Gemma 2 models (9B and 27B)github.com · 2024-06-28
  13. ArticleShow HN: Gemma Scope – Exploring the Inner Workings of Gemma 2 2Bneuronpedia.org · 2024-07-31
  14. ArticleGoogle announces Gemma 2, a 27B-parameter version of its open modeltechcrunch.com · 2024-05-15

Where to run it

Live hosted endpoints, cheapest first, USD per million tokens. Ties break on output price, then provider name.

#Cheapest 2In /1MOut /1MContextQuantSeen
1 NextBit $0.65 $0.65 8K int4 Oct 24
2 Featherless — — — — Oct 25

All 2 providers →

Where it's used

Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.

Served by 2 API providers; 14 quantized builds for local use.

  • As a service2Featherless, NextBit
  • On own hardware14quantized builds
  • Built on3728 fine-tunes · 9 adapters · 0 merges
  • In apps20Hugging Face Spaces

Top Spaces running it: phyloforfun/VoucherVision ♥ 13 · artificialguybr/CanIRunThisLLM ♥ 9 · Ffftdtd5dtft/gfgf ♥ 3

Will it fit?

Weights only, an estimate: leave room for the KV cache. Method

Fits a 64 GB Mac (unified memory) at INT4

About 16 GB at INT4, 33 GB at FP8, 65 GB at BF16: weights only, so leave room for the KV cache.

16 GB card64 GB Mac2×80 GB8×80 GBINT4: about 16 GB of weights (estimate; KV cache and activations extra)INT416 GBFP8: about 33 GB of weights (estimate; KV cache and activations extra)FP833 GBBF16: about 65 GB of weights (estimate; KV cache and activations extra)BF1665 GB
16 GB card64 GB Mac2×80 GB8×80 GBINT4: about 16 GB of weights (estimate; KV cache and activations extra)INT416 GBFP8: about 33 GB of weights (estimate; KV cache and activations extra)FP833 GBBF16: about 65 GB of weights (estimate; KV cache and activations extra)BF1665 GB

Estimate params × bytes per param × 1.2 · how we estimate

The story so far

Weights, every dated article, and repo commits on one rail.

Most coverage came in the first two weeks

Teal is the weights (2024-06-24), amber triangles are 10 dated articles (each links to it), grey ticks are repo commits.

MarAprMayJunJulAugSepOctNovDecJanFebMarAprMayJunJulAugSepOctNovDecJanFebMarAprMayJunJulAugSepOct2024-08-07: 2 repo commits2024-07-02: 2 repo commits2024-06-27: 7 repo commits2024-06-25: 3 repo commits2024-06-24: 3 repo commitsCPUs Aren't Dead. Gemma2B Out Scored GPT-3.5 Turbo on Test That Made It Famous · 2026-04-15Gemma 2: Improving Open Language Models at a Practical Size [pdf] · 2024-06-27Gemma 2B 60 tk/s WebGPU on M1 · 2024-03-07Gemma explained: What's new in Gemma 2 · 2024-08-22New FOSS All-Local AI with Ollama, Gemma 2, and Fabric · 2024-07-012x faster Gemma 2 finetuning and 63% less VRAM · 2024-07-04Show HN: Finetune a Gemma 2B model for codegen · 2024-02-28Gemma 2B with 10M context length · 2024-05-09Gemma 2 on AWS Lambda with Llamafile · 2024-07-06Gemma 2B: Scores better than GPT 3.5 Turbo · 2024-08-01Weights 2024-06-24 · 2024-06-24Weights 2024-06-24now · 2 providers
MarAprMayJunJulAugSepOctNovDecJanFebMarAprMayJunJulAugSepOctNovDecJanFebMarAprMayJunJulAugSepOct2024-08-07: 2 repo commits2024-07-02: 2 repo commits2024-06-27: 7 repo commits2024-06-25: 3 repo commits2024-06-24: 3 repo commitsCPUs Aren't Dead. Gemma2B Out Scored GPT-3.5 Turbo on Test That Made It Famous · 2026-04-15Gemma 2: Improving Open Language Models at a Practical Size [pdf] · 2024-06-27Gemma 2B 60 tk/s WebGPU on M1 · 2024-03-07Gemma explained: What's new in Gemma 2 · 2024-08-22New FOSS All-Local AI with Ollama, Gemma 2, and Fabric · 2024-07-012x faster Gemma 2 finetuning and 63% less VRAM · 2024-07-04Show HN: Finetune a Gemma 2B model for codegen · 2024-02-28Gemma 2B with 10M context length · 2024-05-09Gemma 2 on AWS Lambda with Llamafile · 2024-07-06Gemma 2B: Scores better than GPT 3.5 Turbo · 2024-08-01Weights 2024-06-24 · 2024-06-24Weights 2024-06-24now · 2 providers

What people built on it

51 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.

51 derivatives on the Hub, most of them fine-tunes

From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.

14 quantized (Hugging Face model tree)28 fine-tunes (Hugging Face model tree)9 adapters (Hugging Face model tree)14 quantized28 fine-tunes9 adapters
14 quantized (Hugging Face model tree)28 fine-tunes (Hugging Face model tree)9 adapters (Hugging Face model tree)14 quantized28 fine-tunes9 adapters
GGUF mradermacher/gemma-2-27b-i1-GGUF for llama.cpp, Ollama, LM Studio 133.5k downloads MLX mlx-community/gemma-2-27b-bf16-4bit for Apple silicon 400 downloads

All variants →

Reference

Everything else, folded: the full tables, every source.

Benchmarks

13 rows

Benchmarks 13 rows

measured third-party leaderboard rowauthor-reported number from the model card or eval files (13)No confidence intervals unless the source gave them; settings differ between rows, so compare within a benchmark, not across. Date = the benchmark's own date when it gives one, else when we read it.
BenchmarkValueSetting / notesReported byDate
author-reported ARC Challenge · Chat and follow instructions71.425-shotmodel author2026-10-07
author-reported Commonsense QA sets · Chat and follow instructions83.20-shotmodel author2026-10-07
author-reported Commonsense QA sets · Chat and follow instructions53.70-shotmodel author2026-10-07
author-reported Commonsense QA sets · Chat and follow instructions84.80-shotmodel author2026-10-07
author-reported Commonsense QA sets · Chat and follow instructions88.60-shotmodel author2026-10-07
author-reported Commonsense QA sets · Chat and follow instructions83.75-shotmodel author2026-10-07
author-reported Commonsense QA sets · Chat and follow instructions34.55-shotmodel author2026-10-07
author-reported GSM8K · Reason through hard problems745-shot, maj@1model author2026-10-07
author-reported HellaSwag · Chat and follow instructions86.410-shotmodel author2026-10-07
author-reported HumanEval · Write code with agents51.8 pass@1pass@1model author2026-10-07
author-reported MBPP · Write code with agents62.63-shotmodel author2026-10-07
author-reported MMLU · Reason through hard problems75.25-shot, top-1model author2026-10-07
author-reported Winogrande · Chat and follow instructions83.7partial scoremodel author2026-10-07

What the card says

1 cited sentence

Summary generated, cited

Gemma 2 27B is a text-to-text, decoder-only large language model with 27.2B parameters.[1][2]

  1. “text-to-text, decoder-only large language models”
  2. parameters human

Generated 2026-10-07 15:13 UTC by parasail-gemma-4-26b-a4b-it; each numeral points at the fact it rests on. Not a verdict.

Use-case evidence

3 use cases

What it is good at 3 use cases

Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.

Facts

21 fields

Listed in LLMs

Facts

Pipeline tagtext-generation
Librarytransformers
ArchitecturesGemma2ForCausalLM
Model typegemma2
Dtypefloat32
Params by dtypeF32: 27,227,128,320
Size bandL
MoEno
Context length8,192 (config.json max_position_embeddings)
Gatedmanual
PapersarXiv:2009.03300, arXiv:1905.07830, arXiv:1911.11641, arXiv:1904.09728, arXiv:1905.10044, arXiv:1907.10641, arXiv:1811.00937, arXiv:1809.02789, arXiv:1911.01547, arXiv:1705.03551, arXiv:2107.03374, arXiv:2108.07732, arXiv:2110.14168, arXiv:2009.11462, arXiv:2101.11718, arXiv:2110.08193, arXiv:1804.09301, arXiv:2109.07958, arXiv:1804.06876, arXiv:2103.03874, arXiv:2304.06364, arXiv:2206.04615, arXiv:2203.09509
LicenseGemma Terms of Use (gemma)
Hub snapshot2026-10-25

Servability

vLLM · SGLang

Servability

vLLM: supported · Gemma2ForCausalLM · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · Gemma2ForCausalLM · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed

vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.

Hardware fit

3 estimates

Hardware fit ESTIMATE

Weights-only memory for gemma-2-27b: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.

BF16 (2 bytes/param)65 GB · Single high-end GPU (<80 GB)
FP8 (1 byte/param)33 GB · Single high-end GPU (<80 GB)
INT4 (0.5 bytes/param)16 GB · Single GPU (<24 GB)

Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.

All variants

40 variants

Lineage

Original model Family: google/gemma-2 51 derivatives on the Hub

quantized: 14finetune: 28adapter: 9

Top variants (40 shown, ranked by downloads)

Relation
Quant
40 rows
Variant Relation Quant Params Band License Author Likes Downloads
mradermacher/gemma-2-27b-i1-GGUF quantized GGUF — — gemma mradermacher 2 133.5k
mradermacher/gemma-2-27b-GGUF quantized GGUF — — gemma mradermacher 4 86.6k
vandijklab/C2S-Scale-Gemma-2-27B finetune — 27.2B L cc-by-4.0 vandijklab 177 25.4k
Mungert/C2S-Scale-Gemma-2-27B-GGUF quantized GGUF — — cc-by-4.0 Mungert 0 19.1k
INSAIT-Institute/BgGPT-Gemma-2-27B-IT-v1.0 finetune — 27.2B L gemma INSAIT-Institute 17 5.4k
NikolayKozloff/gemma-2-27b-Q2_K-GGUF quantized Q2_K — — gemma NikolayKozloff 1 4.4k
NikolayKozloff/gemma-2-27b-Q3_K_S-GGUF quantized GGUF — — gemma NikolayKozloff 1 4.2k
kajuma/gemma-2-27b-instruct quantized GGUF 27.2B L gemma kajuma 0 2.6k
zeromodels/gemma-2-27b finetune — — — gemma zeromodels 0 1.9k
KINS-ai/ATOMIS finetune — 32.9B L gemma KINS-ai 9 1.1k
Aratako/Llama-Gemma-2-27b-SFT-trial1 finetune — 27.2B L llama3.1gemma Aratako 1 1.4k
Dido599999/MD1RoBoTs-Gemma2-27B finetune — 27.2B L gemma Dido599999 0 1.2k
BBQGOD/DeepSeek-GRM-27B finetune — 28.4B L gemma BBQGOD 4 594
cjvt/GaMS-27B finetune — 27.2B L gemma cjvt 2 555
yabs/gemma-2-27b-Q4_K_M-GGUF quantized GGUF — — gemma yabs 1 524
mlx-community/gemma-2-27b-bf16-4bit quantized -bf16 27.2B L gemma mlx-community 0 400
itlwas/gemma-2-27b-Q4_K_M-GGUF quantized GGUF — — gemma itlwas 0 375
mlx-community/gemma-2-27b-bf16-8bit quantized -bf16 27.2B L gemma mlx-community 0 367
BBQGOD/DeepSeek-GRM-27B-MetaRM finetune — 27.2B L gemma BBQGOD 3 188
ai-team-20/gemma-3-27b-2 finetune — 27.2B L gemma ai-team-20 0 284
Hibiki252/gemma-2-27b-4bit quantized -4bit 28.0B L apache-2.0 Hibiki252 1 203
NAPS-ai/naps-gemma-2-27b-v0.1.0 finetune — 27.2B L apache-2.0 NAPS-ai 0 248
cuticle999/gemma-2-27b-gguf quantized gguf — — — cuticle999 0 226
NAPS-ai/naps-gemma-2-27b-v-0.1.0 finetune — 27.2B L apache-2.0 NAPS-ai 0 222
azuki-digital/gemma-2-27b-it-2_lora_merged finetune — 27.2B L gemma azuki-digital 0 202
ai-team-20/gemma-3-27b-1 finetune — 27.2B L gemma ai-team-20 0 177
lmg-anon/vntl-gemma2-27b-qlora adapter — — — gemma lmg-anon 0 152
ernie-research/TLDR-Gemma-2-27B-MA-PPO-Fixed5 finetune — 27.2B L mit ernie-research 1 67
Ffftdtd5dtft/gemma-2-27b-Q2_K-GGUF quantized Q2_K — — gemma Ffftdtd5dtft 1 64
shizwick/google-gemma-2-27b_causal-lm adapter — — — — shizwick 0 104
lmg-anon/vntl-gemma2-27b-hf adapter — 28.4B L gemma lmg-anon 2 0
benzlokzik/gemma-2-27b-quantized-8bit-uint quantized -quant 27.2B L apache-2.0 benzlokzik 0 89
c01zaut/gemma-2-27b-it-rk3588-1.1.1 finetune — — — gemma c01zaut 0 75
c01zaut/gemma-2-27b-it-rk3588-1.1.2 finetune — — — gemma c01zaut 0 70
shizwick/google_gemma-2-27b_guard-retrained adapter — — — — shizwick 0 69
edward-google/C2S-Scale-Gemma-2-27B finetune — 28.4B L cc-by-4.0 edward-google 0 65
YY1128/gemma2-27b-itengen-merged finetune — 27.2B L apache-2.0 YY1128 0 53
Dylan1999/gemma27b_ssl finetune — 27.2B L gemma Dylan1999 0 53
Spice-Tateda/gemma-2-27b-cp finetune — 28.4B L apache-2.0 Spice-Tateda 0 40
shizwick/google-gemma-2-27b_causal-lm-rank128 adapter — — — — shizwick 0 34

Repository changelog

17 commits

Repository changelog 17 commits

DateCommitSHA
2024-08-07 Update tokenizer files (#14) 938270f
2024-08-07 Update README.md (#13) 457162e
2024-07-02 Update README.md (#10) 4333076
2024-07-02 Default attention to eager implementation (#12) 5836c3a
2024-06-27 Fix link (#7) e21d016
2024-06-27 Link to terms (#6) c42c091
2024-06-27 Remove experimental notice (#5) 00483a6
2024-06-27 Update README.md (#4) f1e97a5
2024-06-27 Update wheel 66307a0
2024-06-27 Update README.md (#3) da4e4e3
2024-06-27 Update wheel 21bbe45
2024-06-25 Update checkpoint (#2) 995576f
2024-06-25 Tokenizer and transformers wheel (#1) 5028fc8
2024-06-25 Upload folder using huggingface_hub e79e5a0
2024-06-24 Upload folder using huggingface_hub 6323ee6
2024-06-24 Create README.md 79d56dc
2024-06-24 initial commit 76e0d2c

All providers

2 providers · 2 endpoints

Where to run 2 providers · 2 endpoints

Hosted endpoints observed for gemma-2-27b, sorted by input price (USD per million tokens). Each row names its observer and date; prices change daily, so check the source before quoting. Cheapest observed input price: $0.65/M.

ProviderModel refContextQuant$/M in$/M outtok/sTTFTToolsStatusObserved
NextBit
openrouter
 google/gemma-2-27b-it 8K int4 $0.65 $0.65 — — — live 2026-10-24
Featherless
hf
 google/gemma-2-27b — — — — — — — live 2026-10-25

Articles and coverage

16 pieces

Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency. Summaries are generated from a stored copy of the article text; every sentence cites a span of that text, and nothing is shown that failed the citation check.

Official2

Gemma 2B 60 tk/s WebGPU on M1
mediapipe-studio.webapps.google.com2024-03-07officialabout the gemma 2 familyHN 30 pts · 15 comments23 words
Gemma explained: What's new in Gemma 2
developers.googleblog.com2024-08-22officialauthorityabout the gemma 2 familyHN 14 pts · 0 comments891 words
Google released Gemma 2, an open model suite available in 2B, 9B, and 27B parameter sizes that introduces architectural innovations like Alternating Local and Global Attention and Logit Soft-Capping. The 27B variant features 46 layers and a 256000 vocabulary, utilizing Grouped-Query Attention to improve efficiency alongside RMSNorm for Pre and Post-Normalization for training stability. The 2B model demonstrates edge-device capability by outperforming all GPT-3.5 models on the Chatbot Arena.written, reviewed, cited
  • The 2B model uses 4 heads for k_proj and v_proj, 8 for q_proj and o_proj with a head size of 256.
  • Gemma 2 2B outperforms all GPT-3.5 models on the Chatbot Arena at a size runnable on edge devices.
  • Replacing MHA with GQA results in comparable performance while offering parameter efficiency and faster inference times.

Analysis6

CPUs Aren't Dead. Gemma2B Out Scored GPT-3.5 Turbo on Test That Made It Famous
seqpu.com2026-04-15HN front pageabout the gemma 2 familyHN 100 pts · 52 comments
Gemma 2: Improving Open Language Models at a Practical Size [pdf]
storage.googleapis.com2024-06-27HN front pageabout the gemma 2 familyHN 328 pts · 172 comments9,878 words
Google DeepMind releases Gemma 2, a family of open language models ranging from 2 billion to 27 billion parameters. The 27B model trains on 13 trillion tokens of primarily-English data, while the 2B and 9B models use knowledge distillation to simulate training beyond 50× the compute-optimal quantity. All models inherit a vocabulary with 256k entries designed for multilingual use.written, reviewed, cited
  • The 2B and 9B models are trained with knowledge distillation instead of next token prediction.
  • Gemma 2 models use interleaved local sliding window and global attention layers in every other layer.
  • The 27B model requires 6144 TPUv5p chips for training with 768-way data replication and 8-way model sharding.
2x faster Gemma 2 finetuning and 63% less VRAM
unsloth.ai2024-07-04authorityabout the gemma 2 familyHN 3 pts · 1 comments
Show HN: Finetune a Gemma 2B model for codegen
github.com2024-02-28authorityabout the gemma 2 familyHN 5 pts · 2 comments246 words
Gemma 2B with 10M context length
github.com2024-05-09authorityabout the gemma 2 familyHN 6 pts · 1 comments
Gemma 2 on AWS Lambda with Llamafile
unremarkable.ai2024-07-06about the gemma 2 familyHN 10 pts · 3 comments

Community2

New FOSS All-Local AI with Ollama, Gemma 2, and Fabric
news.ycombinator.com2024-07-01about the gemma 2 familyHN 33 pts · 6 comments301 words
RAGAgent is a new FOSS all-local AI project where chat history is vector searched for context assembly and Fabric is used for system and user messages. Unlike OpenWebUI, it is a real agent rather than just a frontend for default REPLs. The Gemma 2 27B model was broken, but a fix is already committed and underway to the pip repo.written, reviewed, cited
Gemma 2B: Scores better than GPT 3.5 Turbo
twitter.com2024-08-01about the gemma 2 familyHN 8 pts · 0 comments
All coverage tracked for this model (16) — append-only, nothing is ever removed
TitleSourcePublishedFirst seenSignals
CPUs Aren't Dead. Gemma2B Out Scored GPT-3.5 Turbo on Test That Made It Famousseqpu.com2026-04-152026-10-07hn
Gemma explained: What's new in Gemma 2developers.googleblog.com official2024-08-222026-10-07hn
Colab Notebook for finetuning Gemma 2 2B IT modelcolab.research.google.com official2024-08-032026-10-07hn
Google releases Gemma 2 2B, ShieldGemma and Gemma Scopehuggingface.co2024-08-012026-10-07hn
Gemma 2B: Scores better than GPT 3.5 Turbotwitter.com2024-08-012026-10-07hn
Run Google Gemma 2 2B 100% Locallynews.ycombinator.com2024-07-312026-10-07hn
Show HN: Gemma Scope – Exploring the Inner Workings of Gemma 2 2Bneuronpedia.org2024-07-312026-10-07hn
Gemma 2 on AWS Lambda with Llamafileunremarkable.ai2024-07-062026-10-07hn
2x faster Gemma 2 finetuning and 63% less VRAMunsloth.ai2024-07-042026-10-07hn
New FOSS All-Local AI with Ollama, Gemma 2, and Fabricnews.ycombinator.com2024-07-012026-10-07hn
Ollama v0.1.47 release. Added support for Google Gemma 2 models (9B and 27B)github.com2024-06-282026-10-07hn
Gemma 2: Improving Open Language Models at a Practical Size [pdf]storage.googleapis.com2024-06-272026-10-07hn
Google announces Gemma 2, a 27B-parameter version of its open modeltechcrunch.com2024-05-152026-10-07hn
Gemma 2B with 10M context lengthgithub.com2024-05-092026-10-07hn
Gemma 2B 60 tk/s WebGPU on M1mediapipe-studio.webapps.google.com official2024-03-072026-10-07hn
Show HN: Finetune a Gemma 2B model for codegengithub.com2024-02-282026-10-07hn

Papers

8 papers

Papers 8

arXiv links found in the model card and its tags. "Peer-reviewed" is shown only with evidence: a published venue in OpenAlex (measured) or the authors' own journal reference on arXiv (reported).

Similar models

3 models

Similar models

  • medgemma-27b-text-it · 27.0B
    same size band (L), same task (text-generation)
  • Qwen3-32B · 32.8B
    same size band (L), same task (text-generation), more providers, permissive license
  • Qwen2.5-Coder-32B-Instruct · 32.8B
    same size band (L), same task (text-generation), more providers, permissive license

Sources

3 sources

Sources

Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:41 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:13 UTC · Last verified 2026-10-25

Follow gemma-2-27b

Watch this model

Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.